Added benchmarks for new implementations

Author: Andrey Oskin

benchmark/bench01_distance.jl

@@ -17,12 +17,6 @@ centroids = rand(10, 2)
 d = Vector{Float64}(undef, 100_000)
 suite["100kx10"] = @benchmarkable ParallelKMeans.colwise!($d, $X, $centroids)
 
-# for reference
-metric = SqEuclidean()
-#suite["100kx10_distances"] = @benchmarkable Distances.colwise!($d, $metric, $X, $centroids)
-dist = Distances.pairwise(metric, X, centroids, dims = 2)
-min = minimum(dist, dims=2)
-suite["100kx10_distances"] = @benchmarkable $d = min
 end # module
 
 BenchDistance.suite

benchmark/bench02_kmeans.jl

@@ -0,0 +1,28 @@
+module BenchKMeans
+using Random
+using ParallelKMeans
+using BenchmarkTools
+
+suite = BenchmarkGroup()
+
+Random.seed!(2020)
+X = rand(10, 100_000)
+
+centroids3 = ParallelKMeans.smart_init(X, 3, 1, init="kmeans++").centroids
+centroids10 = ParallelKMeans.smart_init(X, 10, 1, init="kmeans++").centroids
+
+suite["10x100_000x3x1     Lloyd"] = @benchmarkable kmeans($X, 3, init = $centroids3, n_threads = 1, verbose = false, tol = 1e-6, max_iters = 1000)
+suite["10x100_000x3x1  Hammerly"] = @benchmarkable kmeans(Hamerly(), $X, 3, init = $centroids3, n_threads = 1, verbose = false, tol = 1e-6, max_iters = 1000)
+
+suite["10x100_000x3x2     Lloyd"] = @benchmarkable kmeans($X, 3, init = $centroids3, n_threads = 2, verbose = false, tol = 1e-6, max_iters = 1000)
+suite["10x100_000x3x2  Hammerly"] = @benchmarkable kmeans(Hamerly(), $X, 3, init = $centroids3, n_threads = 2, verbose = false, tol = 1e-6, max_iters = 1000)
+
+suite["10x100_000x10x1    Lloyd"] = @benchmarkable kmeans($X, 10, init = $centroids10, n_threads = 1, verbose = false, tol = 1e-6, max_iters = 1000)
+suite["10x100_000x10x1 Hammerly"] = @benchmarkable kmeans(Hamerly(), $X, 10, init = $centroids10, n_threads = 1, verbose = false, tol = 1e-6, max_iters = 1000)
+
+suite["10x100_000x10x2    Lloyd"] = @benchmarkable kmeans($X, 10, init = $centroids10, n_threads = 2, verbose = false, tol = 1e-6, max_iters = 1000)
+suite["10x100_000x10x2 Hammerly"] = @benchmarkable kmeans(Hamerly(), $X, 10, init = $centroids10, n_threads = 2, verbose = false, tol = 1e-6, max_iters = 1000)
+
+end # module
+
+BenchKMeans.suite

src/hamerly.jl

@@ -303,11 +303,11 @@ function chunk_update_bounds!(containers, r, r1, r2, pr1, pr2)
 
     @inbounds for i in r
         label = labels[i]
-        ub[i] += 2*sqrt(ub[i] * p[label]) + p[label]
+        ub[i] += 2*sqrt(abs(ub[i] * p[label])) + p[label]
         if r1 == label
-            lb[i] += pr2 - 2*sqrt(pr2*lb[i])
+            lb[i] += pr2 - 2*sqrt(abs(pr2*lb[i]))
         else
-            lb[i] += pr1 - 2*sqrt(pr1*lb[i])
+            lb[i] += pr1 - 2*sqrt(abs(pr1*lb[i]))
         end
     end
 end